Attention please!
THE INFORMATION GIVEN IN THIS APPLICATION NOTE IS GIVEN AS A HINT FOR THE IMPLEMENTATION OF THE INFINEON TECHNOLOGIES COMPONENT ONLY AND SHALL NOT BE REGARDED
AS ANY DESCRIPTION OR WARRANTY OF A CERTAIN FUNCTIONALITY, CONDITION OR QUALITY
OF THE INFINEON TECHNOLOGIES COMPONENT. THE RECIPIENT OF THIS APPLICATION NOTE
MUST VERIFY ANY FUNCTION DESCRIBED HEREIN IN THE REAL APPLICATION. INFINEON
TECHNOLOGIES HEREBY DISCLAIMS ANY AND ALL WARRANTIES AND LIABILITIES OF ANY KIND
(INCLUDING WITHOUT LIMITATION WARRANTIES OF NONINFRINGEMENT OF INTELLECTUAL
PROPERTY RIGHTS OF ANY THIRD PARTY) WITH RESPECT TO ANY AND ALL INFORMATION
GIVEN IN THIS APPLICATION NOTE.
Information
For further information on technology, delivery terms and conditions and prices please contact your
nearest Infineon Technologies Office (www.infineon.com).
Warnings
Due to technical requirements components may contain dangerous substances. For information on the
types in question please contact your nearest Infineon Technologies Office. Infineon Technologies
Components may only be used in lifesupport devices or systems with the express written approval of
Infineon Technologies, if a failure of such components can reasonably be expected to cause the failure of
that lifesupport device or system, or to affect the safety or effectiveness of that device or system. Life
support devices or systems are intended to be implanted in the human body, or to support and/or maintain
and sustain and/or protect human life. If they fail, it is reasonable to assume that the health of the user or
other persons may be endangered.
AN 201301
Subjects: Design of Phase Shifted FullBridge Converter with Current Doubler Rectifier
Author: Sam AbdelRahman (IFNA PMM SMD AMR PMD 2)
We Listen to Your Comments
Any information within this document that you feel is wrong, unclear or missing at all? Your feedback will
help us to continuously improve the quality of this document. Please send your proposal (including a
reference to this document) to: [Sam.AbdelRahman@infineon.com]
Table of contents
1 Introduction .................................................................................................................................................. 4
2 FullBridge Converter with Current Doubler Rectifier ............................................................................. 4
3 Modes of Operation ..................................................................................................................................... 5
4 Energy and Deadtime Conditions for Acheiving ZVS .............................................................................. 8
5 Design Equations and Power Losses .....................................................................................................10
6 References .................................................................................................................................................19
Introduction
While the frontend stage of an AC/DC rectifier achieves power factor correction and regulates the bus voltage
to a dc value (~390V), the dcdc stage must step down the bus voltage and provide a galvanically isolated and
tightly regulated dc output (eg. 12V, 24V, 48V). This document is intended to discuss the topology and
operation of the dcdc stage for medium to high power applications ( >400W), and provide detailed design
equations with examples.
A wide range of isolated topologies are available for the dcdc stage, but the choice depends primarily on
power level, complexity and cost. Ideally it is desired to select the topology with the least cost and complexity,
nevertheless, one that can handle the power level with reliable and acceptable performance.
Figure 1.1 shows a chart for topology selection, it must be noted that topologies in this chart are not necessarily
limited to the indicated power ranges, the intention of the chart is only to illustrate topologies power handling
and their common application, relatively.
Figure 1.1
According to specifications in Table 1 and the chart above, 600W could be realized with either a halfbridge or
fullbridge. 600W falls in the high end of the halfbridge power handling range, while a fullbridge can handle
that power with less stress and better performance. A fullbridge has half the rms current compared to a halfbridge, also, it can be implemented with phase shift control which provides Zero Voltage Switching (ZVS) for
primary side switches.
Since the output is 12V and 50A, current double rectifier with synchronous rectification (Figure 2.1) is the most
suitable for such high current application, as it splits the output current between two filter inductors, which
reduces conduction losses, improves thermal distribution, and allows for lower profile, in addition to the ripple
cancellation effect on the output capacitance.
Table 1 Specifications
Input voltage
390 V
Output voltage
12 V
Maximum power
600 W
Switching frequency
200 kHz
20%
12 mVpp
DC Bus
DC/DC
Converter
PFC
Converter
VacAC
Load
+
Vo

IL1
DC Bus
A
L1
+
C
VA IA

SR2
Ip
IC
Lk
+
Vp

+ VL1 ISR2
+
Vs

+
Vo

IL2
L2
SR1
+ VL2 ISR1
Figure 2.1
Modes of Operation
Figure 3.1 shows the equivalent circuit of each mode and key waveforms, switches A and B are switched
complimentary with 50% duty cycle minus a short dead time, switches C and D are also switched
complimentary with 50% duty cycle minus a short dead time, phase shift control between the two switches pairs
A,B and C,D is used for output voltage regulation.
Mode 1: (t0t1)
Duty Cycle Loss Mode
At t=t0, switch A is turned on with ZVS (as a consequence of Mode 10 below), transformer secondary voltage Vs
will remain zero and both secondary side rectifiers SR1 and SR2 will remain conducting current eventhough
that SR2 is gated off, both inductors L1 and L2 are discharging, secondary voltage Vs remain zero until the
primary current Ip reverses its direction and rise to reach the reflected output inductor current I L1*Ns/Np at t=t1. Ip
rises with a slope as the input voltage Vin charges the leakage inductor Lk, and SR2 current slopes down to
reach zero at t=t1. No power is delivered to the output in this mode.
Mode 2: (t1t2)
Power Delivery Mode
At t=t1, the transformer secondary voltage Vs is equal to Vin*Ns/Np, the output inductor L1 is charging, the output
inductor L2 is discharging, SR1 carries both inductors current. The effective phase shift Pheff starts in this mode.
The primary winding current Ip is equal to the reflected output inductor current IL1*Ns/Np.
Mode 3: (t2t3)
Switch C ZVS Mode
At t=t2, switch D is turned off, the primary current Ip charges the capacitance of switch D and discharges the
capacitance of switch C. when switch C is discharged to zero, its body diode conducts to achieve zero voltage
switching condition, the transformer secondary voltage Vs becomes zero and both SRs carry current.
Mode 4: (t3t4)
Freewheeling Mode
At t=t3, switch C is turned on with ZVS, current Ip freewheels through switches A and C, the transformer
secondary voltage Vs is zero, inductor L1 discharges by the output voltage through SR2, and inductor L 2
discharges through SR1.
Mode 5: (t4t5)
Switch B ZVS Mode
At t=t4, switch A is turned off, the primary current charges the capacitance of switch A and discharges the
capacitance of switch B. when switch B is discharged to zero, its body diode conducts to achieve zero voltage
switching condition, the transformer secondary voltage Vs remains zero and both SRs keeps conducting.
Mode 6: (t5t6)
Duty Cycle Loss Mode
At t=t5, switch B is turned on with ZVS, transformer secondary voltage Vs will remain zero and both secondary
side rectifiers SR1 and SR2 will remain conducting current eventhough that SR1 is gated off, Bothe inductors L 1
and L2 are discharging. secondary voltage Vs remain zero until the current Ip reverses its direction and rise (in
the negative direction) to reach the reflected output inductor current IL2*Ns/Np at t=t6. Ip changes with a slope as
the input voltage Vin discharges the leakage inductor Lk, and SR1 current slopes down to reach zero at t=t 6. No
power is delivered to the output in this mode.
Mode 7: (t6t7)
Power Delivery Mode
At t=t6, the transformer secondary voltage Vs is equal to Vin*Ns/Np, the output inductor L2 is charging, the output
inductor L1 is discharging, SR2 carries both inductors current. The effective phase shift Pheff starts again in this
mode. The primary current Ip (in the negative direction) is equal to the reflected output inductor current
IL2*Ns/Np.
Mode 8: (t7t8)
Switch D ZVS Mode
At t=t7, switch C is turned off, the primary current Ip charges the capacitance of switch C and discharges the
capacitance of switch D. when switch D is discharged to zero, its body diode conducts to achieve zero voltage
switching condition, the transformer secondary voltage Vs becomes zero and both SRs carry current.
Mode 9: (t8t9)
Freewheeling Mode
At t=t8, switch D is turned on with ZVS, current Ip freewheels through switches B and D, the transformer
secondary voltage Vs is zero, inductor L1 discharges by the output voltage through SR2, and inductor L 2
discharges through SR1.
SR2
Lk
+
Vs

+
Vp
B
L1
+ VL1 
+
Vo

SR2
+ VL1 
SR1
+ VL2 
Lk
+ VL2 
SR1
+
Vs

+
Vp

L2
T=1/f
T/2deadtime
Mode 2: t1t2
Switch C ZVS Mode
L1
+ VL1 
SR2
Lk
+
Vp

+
Vs

L1
+
Vo

+ VL2 
SR1
+ VL1 
SR1
+ VL2 
+
Vs

+
Vp

L2
SR2
Lk
+
Vo

L2
Mode 3: t2t3
Freewheeling Mode
+ VL1 
SR1
+ VL2 
+
Vs

+
Vo

Vin
Vp
Vs
L2
A
B
C
D
SR1
SR2
L1
SR2
Lk
+
Vp

Pheff*T
Ph*T
T/2
L2
Mode 1: t0t1
Vin*Ns/Np
Vin
Vin*Ns/Np
Mode 4: t3t4
VA
+ VL1 
SR2
Lk
+
Vp

+
Vs

+
Vo

L1
C
+
Vp

L2
B
+ VL2 
SR1
SR2
+ VL1 
Lk
+
Vs

Vin*Ns/Np  Vo
+
Vo

I_L1min
+ VL2 
+ VL1 
SR2
Lk
+
Vp

+
Vs

IL1
SR2
+
Vp
B
+
Vs

SR1
L1
+ VL1 
SR2
Lk
+
Vp

+
Vs

+
Vo

A
+
Vo

L2
IA
L1
C
SR2
+ VL1 
Lk
+
Vs

+
Vp
B
+ VL2 
SR1
Freewheeling Mode
C
+
Vo

SR2
+ VL1 
SR1
+ VL2 
+
Vs

IC
L2
SR1
+ VL2 
ISR1
t0
t1
L1
Lk
+
Vp

I_L1max*Ns/Np
 I_L1min*Ns/Np
Ip
+ VL2 
Mode 8: t7t8
A
IL2
Mode 7: t6t7
Switch D ZVS Mode
+ VL1 
L2
Mode 6: t5t6
VL2
L1
Lk
+ VL2 
SR1
+
Vo

L2
Vo
I_L1max
Mode 5: t4t5
A
VL1
L2
SR1
Vin
+
Vo

t2 t4 t6
t3 t5
L2
Mode 9: t8t9
Switch A ZVS Mode
L1
SR2
+ VL1 
Lk
+
Vp
B
+
Vs

A
+
Vo

L1
C
SR1
+ VL2 
+ VL1 
SR1
+ VL2 
Lk
+
Vs

+
Vp

L2
SR2
L2
Figure 3.1
+
Vo

t7 t9
t8 t10
Although Figure 3.1 shows ZVS turn on for all primary side switches, ZVS may be lost partially or completely
due to lack of energy or insufficient deadtime.
During no load and light load conditions, ZVS may be lost if the inductive energy available in the circuit is not
sufficient to charge and discharge the output capacitance of the two FETs in the same bridge leg in addition to
the transformer capacitance.
The Energy condition for achieving ZVS is:
The total capacitive energy in the switching leg and transformer is:
Where
is the transformer capacitance, and
is the MOSFETs equivalent energy related output
capacitance as list in the datasheet, below is an example from the IPW65R310CFD datasheet to show the
parameter and its definition.
The inductive energy available for ZVS for switches C and D is higher than that for switches A and B. For that
reason, switches C and D have a wider ZVS range compared to switches A and B.
In regards to ZVS for both switches C and D, it can be seen in the equivalent circuits of modes 3 and 8 (Figure
3.1) that leakage inductor, magnetizing inductor and reflected inductor all contribute to the ZVS available
energy, therefore, the total inductive energy available for ZVS for switches C and D is:
Where
However, in regards to ZVS for both switches A and B, it can be seen in the equivalent circuits of modes 5 and
10 (Figure 3.1) that the transformer is clamped to zero voltage, and both output inductors currents freewheels in
the SRs, hence the leakage inductor is the only inductive energy source available for ZVS, therefore, the total
inductive energy available for ZVS for both switches A and B is:
Since the magnetizing energy is not a function of the load current, the designer could slightly gap the
transformer core, in order to tune the magnetizing energy such that ZVS range of is extended at light loads
conditions. Another design practice used to extend the ZVS range is to add an external leakage inductor in
series with the transformer primary, which adds to the leakage energy.
Besides having sufficient energy for achieving ZVS, the deadtime must also be enough to make the voltage
transition, which is dependent on the resonant circuit parameters. Figure 4.1 shows a simple illustration of the
ZVS transition during deadtime, it is seent that the voltage (Vds) transitions to zero in a resonant manner with a
resonant frequency equal to f r, therefore the Vds requires at least one forth of the resonant period (Tr/4) to
complete the transition. Deadtime below Tr/4 will cause switching at a nonzero voltage (partial ZVS), on the
other hand, deadtime higher than T r/4 will cause extra body diode conduction losses, furthermore, might cause
losing ZVS if the resonant circuit is highly damped (or lossy).
Vgs1
Vgs2
Ids
Vds
Tr /4
Tr=1/fr
Figure 4.1
Figure 4.2 shows the circuit for the ZVS mode of switch C (Mode 3) and its equivalent resonant circuit, the
equivalent resonant circuit is similar for all other ZVS transition modes (modes 5,8,10), and their resonant
frequency (fr) can be calculated as:
Notice that
is used for time calculation purposes,
is the MOSFETs equivalent time related
output capacitance as list in the datasheet, below is an example from the IPW65R310CFD datasheet to show
the
parameter and its definition.
Coss
Lk
Lk
CXFMR
Ik
2*Coss
CXFMR
Coss
Figure 4.2
If we consider both energy and time ZVS design equations discussed above, we can see that increasing the
leakage inductance can increase the energy available for ZVS, thus extending the ZVS load range, but on the
other hand it has a side effect of decreasing the resonant frequency during voltage transition, thus increased
deadtime is required, which is not desired in high switching frequency applications. For that reason, it is logical
to reduce the capacitive energy in the circuit rather than increasing the inductive energy, this implies the
necessity of low MOSFETs output capacitcance for this converter and for other ZVS topologies in general.
The following are design equations for the fullbridge converter with current doubler, also a design example is
integrated to further clarify the usage of all equations.
Since rms currents are calculated for the purpose of estimating power losses, we can neglect the waveforms
details shown in Figure 3.1 such in duty cycle loss modes and ZVS modes, and only consider the power
delivery modes and freewheeling modes. Also since its a CCM operation with a low inductor current ripple
(<50%) we may neglect the current ripple effect on all rms values.
Transformer
Turns ratio of the transformer is designed such that the output voltage regulation is maintained during minimum
input voltage (during the holdup time of the preceding PFC stage), and at full load condition. The duty cycle
loss is a term used to describe the time deducted each half switching cycle to reverse current polarity in the
primary side and commute the current between the secondary side rectifiers, as shown in Figure 3.1, which
reduces the period available for power delivery, therefore, duty cycle loss must be considered when calculation
the turns ratio, otherwise the converter might loose regulation at heavy load conditions especially at cases of
higher leakage inductance values.
Voltage gain is shown in the equation below,
For rms currents, power losses calculation and component selections, We may only consider effective phase
shift pheff (neglect duty cycle loss), as:
Selection of transformer size and shape mainly considers its efficiency and temperature rise, some transformer
design experience along with few iterations are requires to reach an optimized core selection and balanced
core and winding losses. For our design example, we chose core E41/17/123C90.
For acceptable core losses, we limit the maximum magnetic flux to 0.1 Tesla, consequently we can calculate
the number of primary turns.
10
For cores with 3C90 or R material, the following equation can be used to calculate core losses,
Filter Inductor
Inductor current ripple
Filter inductors values L1 and L2 and their peak currents are determined based on the specified inductor current
ripple.
11
Primary MOSFET
In order to select the optimum MOSFET, one must understand the MOSFET requirements in a fullbridge
converter. High voltage MOSFETS have several families based on different technologies, which each target a
specific application, topology or operation. For a ZVS converter such as the fullbridge, the following are some
major MOSFET selection considerations:
Low output capacitance Coss for extended ZVS range and short deadtime designs.
Package selection must consider the resulting total thermal resistance from junction to ambient.
Robust body diode with fast reverse recovery. Although that in normal operation the body diode
current/charge are softly commuted, in some conditions such as startup, load transient, light load or low
leakage inductance, body diode may have hard commutation, it may not have a channel conduction
following its own conduction, or channel conduction might be too short and not enough to completely
sweep out the reverse recovery charge, in such case, as the MOSFET turns off with a high dv/dt while
there are still residual charge in the body diode region, the charge leaving the body diode Pregion may
bias the parasitic npn BJT, causing false turn on and destruction of the MOSFET.
CFD2 CoolMOS series are recommended for ZVS applications in general, and for fullbridge converter in
specific. The low reverse recovery charge Qrr (illustrated in Figure 5.1), time trr and current Irrm of the CFD series
brings a higher margin in repetitive hard commutation of the body diode, which increases the system reliability.
Figure 5.1
12
According to the aforementioned MOSFET selection criteria and to the specification listed in Table 1, MOSFET
IPW65R310CFD is selected, and its parameters will be used for the following calculations.
The MOSFET rms current can be calculated by the following equation, consequently the MOSFET conduction
loss is obtained.
Since its a ZVS converter, turnon loss and output capacitance Coss switching loss are zero.
Turnoff time and loss are:
13
In order to choose the optimum MOSFET in synchronous rectification, the power loss mechanism needs to be
well understood. SRs do not operate as active switches; hence they have no voltagecurrent overlap switching
losses, causing switching losses to be relatively constant across the load range, which makes conduction
losses dominant in the heavy load region. For that reason, SRs are required to have low Ron in order to have a
balanced switching and conduction losses and optimal operation.
Another important portion of the total switching losses is related to the output capacitance C oss and the reverse
recovery charge Qrr of the MOSFET. Considering the turnoff moment, the Qrr has to be removed and the
output capacitance has to be charged up to the secondary side transformer voltage. Due to the very short ontime of the body diode of about 50ns to 100ns, Qrr losses can be neglected, leaving only gate charge loss and
Coss loss as switching losses.
Selection of the optimal SR MOSFET that provides balanced and optimized loss can be according to following
formula:
Where,
If we optimize the operation at mid load point, and apply the formulas above to the 75V OptiMOS 3 technology,
we find that:
Another way for choosing the optimal SR MOSFET is to use the tool shown in Figure 5.2. It is a diagram that is
derived from the optimization equation above with applying the Optimos technology parameters and the SR
operation condition. Note that there will be different diagram for each voltage rating, more details are included
in the Infineon Technologies Application Note: Optimum MOSFET Selection for Synchronous Rectification,
that is listed in the references.
The starting point on this diagram is to determine the transformer voltage (or the SR voltage stress), in our case
35.5V, then we move down to the device we intend use (IPP023NE7N3), then we move left to the switching
frequency (150kHz), then we move up to the SR rms current curve, note that we like optimize the SR selection
at mid load, hence 32.37A/2= 16.19A, since there isnt a line for 16A, we may plot an approximated one
between the 10A and the 20A lines, then we may move to the right to find the optimum Rds(on) value and
further to determine how many parallel FETs are required in order to obtain the optimum Rds(on). For our
example we conclude that a single IPP023NE7N3 MOSFET is enough.
14
Figure 5.2
SR conduction loss
15
Output Capacitor
The output capacitor current ripple is a function othe phase shift, current doubler rectifier is able to fully cancel
the current ripple in output capacitance in the case of Pheff = 0.5.
Output capacitor current ripple, rms current and ESR loss can be calculated by the following equations.
For 12mV capacitor voltage ripple as specified in Table 1, the capacitor value is calculate as:
Input Capacitor
The input capacitor rms current can be calculated by the following equation, consequently the capacitor ESR
loss is obtained.
Note that the above capacitor equations assume that the converter is supplies by a dc supply (input current is
constant). If we consider the overall system architecture, we can see that the input capacitance is preceded
and supplied by the PFC stage, hence the shared capacitance (bus capacitor) has two sources of ripple, the
first is the switching frequency ripple caused by the interaction between the two stages, if the full bridge
switching frequency is synchronized with the PFC stage, then this rms current can be minimized, this is
possible as many control ICs offer the synchronization input to their internal oscillator. The second source of
ripple is the 60 Hz ripple generated by the PFC stage. For better estimation of the rms current, especially in a
nonsynchronized application or if the PFC stage is a variable frequency operation, then simulation might be
the best approach.
16
Heat sink
Multiple MOSFETs can have separate heat sinks or share the same one; however, the selection of the heat
sink is based on its required thermal resistivity.
In case of a separate heat sink for each MOSFET, thermal
resistors are modeled as in Figure 5.3.
PFET1
PFET2
Figure 5.3
In case of a single heat sink for multiple MOSFETs, thermal
resistors are modeled as in Figure 5.4.
The maximum heat sink temperature
allowed is the
minimum outcome of the two equations below
PFET1
PFET2
TJ.FET2
RthJC.FET2
TC.FET2 RthCS.FET2
RthSA
TA
PFET1+PFET2
Figure 5.4
Once
is specified, then the heat sink thermal resistance
can be calculated.
is the thermal resistace from junction to case, this is specified in the MOSFET and Diode datasheets.
is the thermal resistace from case to heatsink, typically low compared to the overall thermal resistance,
its value depends on the the interface material, for example, thermal grease and thermal pad.
is the thermal resistance from heatsink to ambient, this is specified in the heatsink datasheets, it depends
on the heatsink size and design, and is a function of the surroundings, for example, a heatsink could have
difference values for
for different airflow conditions.
is the heatsink temperature,
is the case temperature ,
is the ambient temperature.
is FETs total power loss.
17
Table 2 shows two design examples for the CCM PFC boost converter, for different power levels and switching
frequencies.
Table 2
Design Example #1
Design Example #2
600W 150kHz
1000W 100kHz
E41/17/123C90
E42/21/153C90
33
33
0.089 Tesla
0.112 Tesla
1.139 W
1.622 W
2.273 A
3.788 A
20.55 W
34.281 W
IPW65R310CFD
IPW65R190CFD
Transformer
Core size/material
Filter Inductor
Primary MOSFET
Output Capacitor
Input Capacitor
18
References
[1] Product Brief, 650V CoolMOS CFD2.
http://www.infineon.com/dgdl/PB_CoolMOSCFD2_0004_web.pdf?folderId=db3a30432ea425a4012eb484ea6c29f0&fileId=db3a30432ea425a4012
eb48e250d29f1
[2] CoolMOS Selection Guide.
http://www.infineon.com/dgdl/infineon_CoolMOS_SelectionGuide.pdf?folderId=db3a304314dca389011528372fbb12ac&fileId=db3a30432f91014f012f95fc7c24399
d
[3] Infineon Technologies Application Note: Optimum MOSFET Selection for Synchronous Rectification.
May 2012.
http://www.infineon.com/dgdl/Application+Note++Optimum+Mosfet+Selection+for+Synchronous+Rectification+v2+4.pdf?folderId=db3a3043156fd5730
115939eb6b506db&fileId=db3a30431ff988150120664202606c8b
[4] Infineon Technologies Application Note: CoolMOS CFD2  First 650v Rated Super Junction Mosfet
With Fast Body Diode Suitable For Resonant Topologies. February 2011.
http://www.infineon.com/dgdl/Infineon++Application+Note++650V+CoolMOS+CFD2.pdf?folderId=db3a304314dca389011528372fbb12ac&fileId=db3a30432e25b0
09012e29f7433d382f
[5] Infineon Technologies Article A New 650V Super Junction Device with Rugged Body Diode for Hard
and Soft Switching Applications.
http://www.infineon.com/dgdl/Article_CFD2.pdf?folderId=db3a304314dca38901152836c5a412ab&fileId
=db3a30432d081e66012d9863c18d60f5
19
20